Conference Agenda

The System Analysis Module (SAM) is an advanced system analysis tool under development at Argonne National Laboratory, aiming to provide fast-running, modest-fidelity, whole-plant transient analyses capabilities, which are essential for fast-turnaround design scoping and engineering analyses of advanced reactor concepts. As a MOOSE-based computer code, SAM leverages modern advanced software environments and numerical methods provided by the MOOSE framework, such as its underlying meshing and finite-element library and linear and non-linear solvers. As the computer code is being widely adopted and applied in advanced nuclear reactor analyses, some numerical issues have been revealed that impact the code robustness and execution speed. Such issues could be linked to the usage of continuous Galerkin finite element method (CG-FEM) in solving thermal fluid problems. In this work, we investigate the feasibility of implementing a staggered-grid finite volume method (SG-FVM) in the MOOSE framework to support the development of advanced system analysis codes such as SAM. In this work, we demonstrated that a second-order SG-FVM is successfully implemented under the MOOSE framework as the foundation of a numerical test bed for system analysis code development. The SG-FVM-based code also exhibits superior performance in terms of execution speed based on the results of a suite of selected test problems with different problem sizes and levels of complexity. To further verify the correctness of SG-FVM-based code, it was applied to solve the Protected Loss-Of-Flow (PLOF) transient of the Advanced Burner Test Reactor (ABTR). The results show good agreement with reference results from previous studies.

The thermal-hydraulics codes DRACCAR 2024 and TRACE V5p9 simulated OECD/NEA ISP-53 tests representative of low pressure or low flowrate reflooding scenarios in Loss-of-Coolant-Accidents (LOCA). The ISP-53 tests were selected from the COAL reflooding experiments whose test section mimics PWR fuel assemblies. The objectives of the simulation exercises are to perform code-to-experiment and code-to-code benchmarks through comparisons of the simulations and the experimental results to evaluate the capability of the computer codes for modelling of reflooding thermal-hydraulics under challenging conditions such as low pressure or low flowrate.

Representative experimental results, including a few time-independent and time-dependent parameters, were chosen as figures-of-merit (FoM) to assess each code’s performance. TRACE’s simplified modelling of the rod bundle allows for faster simulations, while DRACCAR’s detailed modelling captures intricate phenomena at the expense of computational cost. The simulation results of both codes exhibited significant deviations from experimental data of the case at low pressure (3bar) and medium flowrate (50kg/m2s), with overestimated quenching speeds and underestimated peak cladding temperatures. The codes’ performance improved for simulation of the case at medium pressure (20bar) and low flowrate (17kg/m2s) although overestimation of quenching speed remained.

The low pressure or low flowrate of reflooding is a challenge for thermal-hydraulics codes to reproduce. Thus, caution should be paid when applying the codes to safety analyses of light water reactors under such conditions. The findings highlight the need for model refinements of thermal-hydraulics codes to address deficiencies in reflooding and quenching predictions, particularly for low pressure scenarios to enhance nuclear reactor safety assessments.

Micro-channel heat exchangers represent a significant innovation in heat transfer technology, offering high thermal efficiency and compact designs potentially suitable for advanced nuclear systems. Despite their potential, limited numerical analyses and experimental results are available in literature that fully characterize their performance, especially under prototypical operating conditions found in nuclear reactors. For this purpose, the Versatile Loop Facility (VLF) was designed and built to test the key components which will be part of the reactor coolant system of the Westinghouse LFR, focusing onthe Primary Heat Exchanger (PHE). The PHE is a hybrid microchannel heat exchanger manufactured using a diffusion bonding process offering a high heat transfer area-to-volume ratio, resulting in exceptional compactness, a significant advantage for the design of the primary reactor pool, as it minimizes required space allocation.

This paper presents a comparative study between RELAP5 and ANSYS-CFX for modeling microchannel heat exchangers, using the PHE of the VLF as a case study with the aim to focus on the temperature distribution, pressure drops and heat transfer coefficients and subsequently to improve the accuracy and reliability of thermal-hydraulic system codes and related modelling methodologies for design assessment and safety analyses.

Results show that the simulations of both computer codes are in a good agreement, meaning that RELAP5 provides a satisfactory overall system-level prediction.

In the reactor Emergency Core Cooling (ECC) scenario of a Pressurized Water Reactor (PWR), the external ECC coolant containing high-concentration boric acid is injected into reactor core to prevent the re-critical. The accurate estimation of the transportation of boric acid in the primary circulation is essential for the system thermal hydraulic code. The annular down-comer is the located up-stream of reactor core entrance, and the flow characteristics within it are significantly more complex than those in conventional piping. The presented work focuses on the accuracy of the boron transportation module of RELAP5 in predicting the transient boron concentration file within the annular down-comer. The experimental data that modeling the ECC scenario is introduced for the applicability evaluation. The simulation model with single-loop channel of annular pipeline component is established first. The simulation result shows the boron concentration inside the single-loop annular pipeline is almost linearly distributed, which deviates significantly from the experimental data. The improved model with four branches of annular pipeline components is proposed, where the lateral nodes of the four branches are interconnected. The improved model is capable of predicting the three-dimensional transportation of boric acid in the annular down-comer. The mean boron concentration in the four azimuthal regions of the experimental model corresponds well with the boron concentration in the corresponding branches. However, neither the conventional model nor the improved model is capable of accounting for the impact of density difference between the ECC coolant and ambient coolant on the transportation of boric acid.

Gas-liquid two-phase flow analyses are heavily involved in evaluating the safety of a nuclear power plant. System analysis codes are used to predict the system behavior of nuclear power plants. The system analysis code does not explicitly treat microscopic thermal-hydraulic behavior but uses constitutive equations incorporating its effects to achieve reliable analysis. The constitutive equations are being improved based on increasingly sophisticated measurement techniques and accumulated knowledge. The Nuclear Regulation Authority (NRA) in Japan has developed the system analysis code AMAGI as a platform to consolidate such state-of-the-art constitutive equations. NRA also continues to make extensive efforts to develop and improve critical constitutive equations, such as the drift-flux model. In this paper, the drift-flux models recently developed by the authors are implemented in AMAGI code, and the performance of the drift-flux models is evaluated by comparing experimental data and calculation results.

An analytical benchmark activity was launched within the OECD/NEA ETHARINUS project to assess the capabilities of system codes to simulate the relevant phenomena associated to the PKL Test J4.2, an Extended Loss of Alternate Power (ELAP) with the activation of the SAfety COndenser (SACO) passive system. The selected experiment allows to analyze the interactions of the primary and secondary systems with the passive system. The activity was divided into two phases: a blind phase and an open phase where participants had a period of time to improve their models. In total, 11 participants took part to the benchmark coming from a broad number of countries and applying different system codes.
The results of the blind phase showed significant discrepancies between the experimental data and the simulation outcomes, as well as among different participants' predictions. This was mainly due to the complexity of the scenario, its long duration, the strong influence of heat losses and the presence of cliff edge phenomena. During the open phase, each participant made specific adjustments to their simulation models. The challenges addressed by each participant were similar and mainly focused on improving the reproduction of heat losses and achieving a better heat balance between the systems.
The ETHARINUS benchmark provided valuable insights into the challenges of simulating complex thermal-hydraulic phenomena in nuclear reactors. The findings and recommendations from this benchmark activity contribute to enhancing the predictive capabilities of system codes and the integration of passive safety systems in nuclear reactors.